Jump to content

Requests for technical support from the VASP team should be posted in the VASP Forum.

ELPH WF CACHE PREFILL: Difference between revisions

From VASP Wiki
Improve ELPH_WF_CACHE_PREFILL: fix TAGDEF/DISPLAYTITLE formatting, expand body, add ELPH_WF_CACHE_MB link, remove ELPH_TRANSPORT
Remove wf cache is full error note
Line 10: Line 10:


Setting <code>ELPH_WF_CACHE_PREFILL = .FALSE.</code> disables pre-filling; wavefunctions are then fetched on demand during the loop. This reduces the upfront communication cost but increases total MPI traffic and is generally slower.
Setting <code>ELPH_WF_CACHE_PREFILL = .FALSE.</code> disables pre-filling; wavefunctions are then fetched on demand during the loop. This reduces the upfront communication cost but increases total MPI traffic and is generally slower.
{{NB|text=Pre-filling requires the cache to be large enough to hold all wavefunctions needed by the calculation. If VASP aborts with <code>wf cache is full</code>, increase {{TAG|ELPH_WF_CACHE_MB}}.}}


==Related tags and articles==
==Related tags and articles==

Revision as of 12:23, 18 June 2026

Template:DISPLAYTITLE ELPH_WF_CACHE_PREFILL = logical
Default: ELPH_WF_CACHE_PREFILL = .TRUE. 

Description: Pre-fills the wavefunction cache before the main electron-phonon loop begins.



When computing electron-phonon matrix elements, VASP caches wavefunctions fetched from remote MPI ranks (see ELPH_WF_CACHE_MB). With ELPH_WF_CACHE_PREFILL = .TRUE. (the default), all required wavefunctions are gathered into the cache in a single communication phase before the main loop starts. This means the loop itself runs with little or no inter-rank MPI traffic.

Setting ELPH_WF_CACHE_PREFILL = .FALSE. disables pre-filling; wavefunctions are then fetched on demand during the loop. This reduces the upfront communication cost but increases total MPI traffic and is generally slower.

Related tags and articles